Spontaneous synaptic potentials were identified at the motor endplate 40 years ago. These were shown to possess amplitudes that could be described by a Gaussian distribution as could the amplitudes of evoked synaptic potentials under conditions of very low probability for secretion. As these Gaussians were identical, the idea of a unit or quantum of transmission was conceived. The failure to obtain similar Gaussian distributions for both spontaneous and low-probability evoked potentials during development of endplates indicated that a unit of transmission was not operating. However both the spontaneous and very low-probability evoked potentials could each be described by mixtures of Gaussians indicating a subunit of transmission might be operative. There are no ganglionic or central synapses at which comparisons have been made between spontaneous and low-probability evoked potentials that show each can be described by a Gaussian distribution, let alone that these are the same indicating a unit of transmission as originally conceived. There is some evidence that mixtures of Gaussians can be used to describe both spontaneous and very low-probability evoked synaptic potential amplitudes, opening up the possibility for a subunit of transmission at these synapses. The vesicle hypothesis, that the quantum of transmission at the endplate is due to the exocytosis of the contents of a synaptic vesicle, was also enunciated nearly 40 years ago. The existence of subunits of transmission has required reconsideration of this hypothesis. Three alternatives are considered: in one, the calcium-transient hypothesis, the subunit of secretion is due to the release of calcium from one of several calcium stores in the nerve terminal, so that several subunits are released when a number of these calcium stores are engaged in a regenerative response to the terminal action potential; a second alternative, the mediatophore hypothesis, is that a subunit of secretion occurs when a single transmitter transport protein channels transmitter across the terminal membrane, several such mediatophore proteins acting in concert then give multiple subunit release; finally, there is the vesicle fusion-pore hypothesis, in which individual transient openings of a fusion-pore channel joining a synaptic vesicle to the terminal membrane are responsible for secretion of a transmitter subunit, with multiple transients giving several subunits. Perhaps we will have distinguished between these possibilities before the quantal hypothesis is 50 years old.
